| Hi-index | 0.00 |
This paper presents a neural network (NN) based model for hysteresis nonlinearity with multivalued mapping. It is proved that the Preisach-type hysteresis can be transformed into a general continuous mapping such as one-to-one or multivalued-to-one mapping, which can be approximated by a universal approximator. The main advantage is that the proposed model is suitable to different working conditions by adjusting the weights of NNs. Finally, the derived model is verified by modelling the behavior of hysteresis involved in a piezoelectric actuator.